3 Unstoppable AI Stocks Owned By Billionaires

Bill Ackman’s Pershing Square held three large AI-linked positions at quarter-end: Amazon at 14.3% of the portfolio, Alphabet at 12.5%, and Meta Platforms at 11.4%, together over 38% of assets. The article argues Amazon’s AWS and custom chips, Alphabet’s Google Cloud and Gemini, and Meta’s AI-enhanced ad targeting are all benefiting from the AI trend. This is primarily an investment commentary piece rather than fresh company-specific news, so market impact is likely limited.

Analysis

The portfolio signal is less about “AI winners” in the abstract and more about where AI monetizes fastest: infrastructure, model distribution, and ad optimization. That favors the hyperscalers and the largest consumer platforms because they can amortize AI capex across huge installed bases, while smaller software names remain exposed to rising inference costs and weaker pricing power. In second-order terms, the real competitive pressure lands on mid-tier cloud providers, standalone ad-tech, and anyone selling generic compute without proprietary workload demand.

The market is still underestimating how AI changes the margin structure of legacy businesses. For AMZN and GOOGL, incremental AI demand can improve utilization of already-built data-center assets, so the near-term earnings lever is not just revenue growth but fixed-cost absorption; that tends to show up over quarters, not days. META is even more interesting because AI does not need to create a new product to matter—if targeting efficiency improves, every basis point of ad yield flows almost directly into operating leverage, which can sustain multiple expansion even in a mature growth profile.

The contrarian risk is that consensus is pricing these as “obvious winners,” which makes valuation and capex the real battleground. If AI demand decelerates, or if capex intensity rises faster than monetization, the stocks can de-rate despite strong fundamentals. The cleaner tell over the next 1-2 quarters is whether cloud backlog, ad conversion, and inference economics continue to improve faster than spend; if not, the trade shifts from multiple expansion to quality-at-a-reasonable-price defense.

The broader read-through is bullish for NVDA in the medium term, even though it is not the primary subject of the positioning signal, because hyperscaler AI spend ultimately feeds the GPU ecosystem and attached networking/storage stack. But the pairwise opportunity is better in the platform names versus the pure-play semiconductor complex if the market starts rotating toward self-funded AI cash generators rather than capex beneficiaries. The setup argues for owning the companies that can both buy the picks-and-shovels and extract the most value from them.

Key Decisions for Investors

• Stay long AMZN / GOOGL / META into the next earnings window; use 3-6 month horizon and size for continued multiple support if AI monetization outpaces capex.
• Pair trade: long META vs short a basket of mid-cap ad-tech names over 1-3 quarters; thesis is AI-driven targeting gains concentrate share in the dominant platform, compressing smaller intermediaries' economics.
• Long AMZN and GOOGL against a short basket of lesser cloud/AI infrastructure providers over 6-12 months; hyperscaler utilization and workload concentration should outcompete commodity compute sellers.
• If positioning becomes crowded, buy 6-9 month downside protection on META or GOOGL rather than outright reducing exposure; valuation risk is the cleanest reversal catalyst, not fundamental collapse.
• Monitor NVDA as a second-order beneficiary and consider relative-value long NVDA vs short high-multiple software names if AI capex stays firm but software monetization lags.